English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Our goal is to understand why the robustness drops after conductingadversarial training for too long. Although this phenomenon is commonlyexplained as overfitting, our analysis suggest that its primary cause isperturbation underfitting. We observe that after training for too long,FGSM-generated perturbations deteriorate into random noise. Intuitively, sinceno parameter updates are made to strengthen the perturbation generator, oncethis process collapses, it could be trapped in such local optima. Also,sophisticating this process could mostly avoid the robustness drop, whichsupports that this phenomenon is caused by underfitting instead of overfitting.In the light of our analyses, we propose APART, an adaptive adversarialtraining framework, which parameterizes perturbation generation andprogressively strengthens them. Shielding perturbations from underfittingunleashes the potential of our framework. In our experiments, APART providescomparable or even better robustness than PGD-10, with only about 1/4 of itscomputational cost.

Overfitting or Underfitting? Understand Robustness Drop in Adversarial Training